Reliability and Validity of a Carbohydrate-Counting Knowledge Questionnaire for Young Australians With Type 1 Diabetes.

OBJECTIVE: To test the reliability and validity of a carbohydrate-counting knowledge questionnaire in young Australians with type 1 diabetes mellitus (T1DM). METHODS: Children or young adults (<20 years) with T1DM, or their parents, completed the 72-item Australian PedCarbQuiz (AusPCQ), adapted from the American PedCarbQuiz, and an expert assessment of carbohydrate-counting knowledge. Responses were scored and summed (0-72, higher scores = greater knowledge). Internal reliability was assessed using Cronbach α, and relative validity using Spearman correlations (with HbA1c) and Bland-Altman analysis (with the expert assessment). RESULTS: Australian PedCarbQuiz reliability (n = 44, mean score = 59.7 ± 5.6) was acceptable (α = 0.83). There was a lack of agreement (mean bias = 10.7, P = 0.008) and significant proportional bias between AusPCQ scores and expert assessments (ß = -0.73 [95% confidence interval, -1.82 to -0.79]; P < 0.001). CONCLUSIONS AND IMPLICATIONS: The AusPCQ was shown to be reliable but not valid in a small sample. Testing in a larger sample is warranted.

Nutritional risk factors in planned oncologic surgery: what clinical and biological parameters should be routinely used?

BACKGROUND: Screening for malnutrition is recommended in hospitalized and planned surgical patients. The aim of this study was to analyze the feasibility and routine prognostic value of using the principal recommended nutritional screening and evaluation tools for cancer patients undergoing major surgery. METHODS: This study is a prospective, 3-month, multicenter observational trial recording weight loss, body mass index, albumin, transthyretin, and PG-SGA. The morbidity rate was assessed on the basis of major complications (MC), whether of an infectious (MIC) or noninfectious (MNIC) nature. RESULTS: Two hundred seventy-five patients were recruited at nine centers. The following percentages were recorded with respect to morbidity: 28.4% MC, 12.7% MIC, and 22.2% MNIC. Univariate analysis revealed a statistical association only between weight loss greater than 10% and MIC and hospital stay. A weight loss of 15% is required to demonstrate an association with either MC, MIC, or MNIC. Body mass index (BMI) was associated only with MNIC, PG-SGA with MC, and albumin <30 g/l was strongly associated with all types of morbidity (MC, MIC, MNIC). Multivariate analysis indicated that only albumin <30 g/l and an operating time of more than 4 h are significantly associated with morbidity. CONCLUSIONS: In this study, the best nutritional factor for detecting the risk of MC is albumin levels below 30 g/l. A weight loss greater than 15% is required to obtain a statistically significant correlation with the existence of MC.